Download - Bell 2006 J Struct Geol
-
8/3/2019 Bell 2006 J Struct Geol
1/17
The internal inclusion trail geometries preserved within a first phase
of porphyroblast growth
T.H. Bell *, M.D. Bruce
School of Earth Sciences, James Cook University, Townsville, Qld 4811, Australia
Received 5 May 2005; received in revised form 1 November 2005; accepted 1 November 2005
Available online 5 December 2005
Abstract
The inclusion trail geometry within a first phase of porphyroblast growth can differ significantly from that preserved by further
enlargement because the porphyroblast forms a rigid mass up against which the rock preferentially strains during ensuing events. The
geometry of the first overgrown inclusion trails is affected by their primary orientation, including any pre-existing curvature, combined with
any heterogeneous rotation of this foliation about the developing stretching lineation. This can impact the apparent timing of foliation
intersection/inflection axes preserved within porphyroblasts (FIAs) that nucleated during the development of a sub-horizontal foliation, but is
readily resolved. 3-D computer analysis of sigmoidal inclusion trails reveals that the asymmetry method for FIA determinations is unaffected
by the cut location relative to the porphyroblast core. Significantly, perfect spiral inclusion trail geometries can be produced from a sigmoidal
shape in cuts up 308 away from the FIA. Therefore, since FIAs in most porphyroblasts bear no relation to matrix structures, there is a 17%
chance that thin-sections cut relative to the foliation lie within 308 of a FIA and could contain such an apparent spiral. FIAs maintain
consistent trends for the first phase of porphyroblast growth accompanying horizontal bulk shortening but may vary in plunge. FIAs have
sub-horizontal plunges for porphyroblasts nucleating during gravitational collapse, but may vary in trend. For all periods of porphyroblast
regrowth the data available indicates that FIAs remain consistently trending and sub-horizontal until the relative direction of plate motion
causing orogenesis changes.
q 2006 Elsevier Ltd. All rights reserved.
Keywords: Porphyroblast nucleation; Porphyroblast growth; Crenulation cleavage; Shear sense
1. Introduction
Over a decade has passed since measurements were first
made on the orientation of foliation intersection/inflection
axes in porphyroblasts (FIAs). For some years now this data
has been providing extended structural histories that have
been lost from the matrix due to repeated shearing or
reactivation of the bedding (e.g. Ham and Bell, 2004),
extended metamorphic PTt paths (e.g. Sayab, 2005),
quantitative correlation of metamorphic and structural data
around oroclines and along and across orogens (e.g. Aerden,
2004), and a means for dating periods of deformation
associated with the successive changes in relative plate
motion that accompany extended orogenesis (e.g. Bell and
Welch, 2002).
One feature of this data has not received much attention.
A first phase of porphyroblast growth does not necessarily
behave like the later phases in terms of the internal inclusion
trail geometry preserved within. This occurs because, once a
porphyroblast has grown, it forms a rigid mass against
which the rock preferentially deforms during ensuing
events, preventing reactivation of the compositional
layering as shown in Fig. 1. Subsequent porphyroblast
growth can entrap the foliations that have formed against the
porphyroblast rim and they maintain predominantly sub-
vertical and sub-horizontal orientations (e.g. Hayward,
1992). However, the first phase of porphyroblast growth
overgrows a crenulation hinge as it begins to develop in a
pre-existing foliation; such crenulations or foliation inflec-
tions may have a large range of orientations depending on
the orientation of this earlier-formed foliation. For example,
if the porphyroblast grows in a deformation event that forms
a sub-vertical foliation, then the resulting FIA will have a
Journal of Structural Geology 28 (2006) 236252
www.elsevier.com/locate/jsg
0191-8141/$ - see front matter q 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsg.2005.11.001
* Corresponding author
E-mail address: [email protected] (T.H. Bell).
http://www.elsevier.com/locate/jsghttp://www.elsevier.com/locate/jsg -
8/3/2019 Bell 2006 J Struct Geol
2/17
trend consistent with those measured from porphyroblaststhat contain earlier formed cores, but a variable plunge
(Fig. 2; Bell and Wang, 1999). If the porphyroblast grows in
a deformation event that forms a sub-horizontal foliation,
then the resulting FIA will have a sub-horizontal plunge but
potentially a trend that is not consistent with those measured
from porphyroblasts with earlier formed cores (Fig. 3). If
there has been a previous crenulation event that was not
accompanied by porphyroblast growth, then other geome-
tries are possible.
To demonstrate the geometric consequences and signifi-
cance of these effects we:
1. Examine the variation in inclusion trail geometry andFIA trend within different porphyroblastic phases in the
same layer, as well as within the same phase in different
layers, from two adjacent samples from the one outcrop;
in these samples there is a remarkable consistency of
FIA trends between porphyroblasts showing multiple
phases of growth and those that grew during the
development of a sub-vertical foliation. This contrasts
dramatically with those that only grew during the
development of a sub-horizontal foliation.
2. Examine the 3-D geometry of sigmoidally shaped
inclusion trails obtained by high-resolution X-ray
Fig. 1. (a) Vertical cross-section through a porphyroblast within an obliquely dipping foliated matrix. (b) Progressive bulk inhomogeneous shortening
involving a component of anticlockwise shear causes antithetic (clockwise shear) on the oblique foliation. The presence of the rigid porphyroblast causes
synthetic clockwise axial plane shear to localize against its margins. (c) Clockwise synthetic shear develops a differentiated foliation against the porphyroblast
margin.
Fig. 2. (a) 3-D sketch showing the location of the 2-D longitudinal section (b) and cross-section (c). Porphyroblasts that grew in crenulation hinges during a
deformation event that formed a sub-vertical foliation SnC1 ((a) and (c)) contain FIAs with the same trend but variable plunges ((a) and (b)).
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 237
-
8/3/2019 Bell 2006 J Struct Geol
3/17
computed tomography (X-ray CT) of a garnet porphyr-
oblast that reveals significant aspects of the variation of
inclusion trails that are possible within a first phase of
porphyroblast growth, which need to be taken into
account when attempting to time porphyroblasts
microstructurally.
3. Discuss the significance and implications of these
phenomena for structural geologists and metamorphic
petrologists working with porphyroblastic rocks.
2. FIA trend variation within samples from the Tommy
Creek Block
2.1. The samples
Samples TC1365 and TC1365i were taken from adjacent
thin pelitic bands in an outcrop of the Tommy Creek Block,
Mount Isa Province of Queensland, Australia (Fig. 4; Lally,
1997). Three different pelitic bands within these samples
contain:
A. garnet porphyroblasts (layer A; sample TC1365i).
B. garnet plus andalusite porphyroblasts (that locally
replace staurolite; layer B; sample TC1365i).
C. garnet and staurolite porphyroblasts (layer C; sample
TC1365).
2.2. Timing of porphyroblast growth and FIA trend
The FIAs preserved in these porphyroblasts were
measured separately for each porphyroblastic phase in
each compositional layer. The foliations and phases of
growth preserved within these porphyroblasts, plus the FIA
trends they define, are as follows:
1. Layer Agarnet porphyroblasts preserve up to six
stages of growth (Fig. 5) that occurred during D2 (S2
sub-vertical, S3 sub-horizontal) through D7 (S7 sub-
horizontal). All have a FIA trending at 3508.
2. Layer Bgarnet porphyroblasts in the andalusite
bearing layer (Fig. 6a) contain sub-vertical S4 with a
clockwise asymmetry (looking NE) into sub-horizontal
S5 and the FIA trends at 458. Local relics of staurolite
porphyroblasts preserved within masses of andalusite
have very similar inclusion trail geometries to garnet,
preserving sub-vertical S4 with a clockwise asymmetry
into sub-horizontal S5 (Fig. 6b). The FIA was not
determined for staurolite because relics of this phase are
only present in a few thin sections, but we expect it to be
similar to that in garnet. Andalusite porphyroblasts grew
early during the development of a matrix crenulation
with a sub-vertical axial plane (S6), and then during the
development of another with a sub-horizontal axialplane (S7); the FIA for andalusite trends at 3508.
3. Layer Cgarnet and staurolite porphyroblasts in layer C
(Fig. 7) contain FIA trending at 350 and 458,
respectively. Garnet grew predominantly during the
development of sub-vertical S4, with some rim growth
during the development of sub-horizontal S5 (Fig. 7c,d),
whereas staurolite grew predominantly during the
development of sub-horizontal S5 (Fig. 7a,b).
2.3. Interpretation
Layer A contains garnet porphyroblasts preserving
multiple phases of growth that accompanied up to six
different deformation events. Once garnet had formed, any
foliations that developed against its rim were preserved by
subsequent phases of growth in the orientation in which they
were produced, as shown in Figs. 5 and 8b. The FIA trend of
3508 is controlled by the strike of the sub-vertical foliation
S2 that predated porphyroblast growth as well as the strike
of S4 and S6 that accompanied it, independent of any
rotational effects or reactivation in the matrix. There was no
rotation of L23 towards the stretching lineation L33, such as
Fig. 3. (a) 3-D sketch showing the location of the 2-D horizontal section (b) and cross-section (c). Porphyroblasts that grew in crenulation hinges during a
deformation event that formed a sub-horizontal foliation SnC1 ((a) and (c)) contain FIAs with the same plunge but variable trends ((a) and (b)).
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252238
-
8/3/2019 Bell 2006 J Struct Geol
4/17
that shown in Fig. 2, prior to porphyroblast nucleation in the
very early stages of D3.
In layer B, garnet only grew during development of sub-
horizontal S5 (Figs. 6a and b and 8d). The regionally
coarsely partitioned nature of D5 (Bell and Hickey, 1998)
caused the L45 intersection lineation to rotate towards the
WE-trending L55 mineral elongation lineation, such as
shown in Fig. 3, at a scale much bigger than a porphyroblast.
Where garnet nucleation accompanied reduction in the scale
of deformation partitioning, it overgrew portions of rock
where L45 had been rotated from 3508 towards the L55
stretching lineation to a trend of 458 (Figs. 3 and 8d).
Andalusite first nucleated in D6, which produced sub-
vertical S6 (Fig. 8e). It then grew during development of a
weak sub-horizontal S7. The 3508 FIA trend in andalusite is
controlled by the 3508 striking sub-vertical S6. This FIA
could vary in plunge from the horizontal if rotation of L56towards the steeply pitching L66 occurred in a similar manner
to that which affected L45 (such as shown in Fig. 3); however,
such rotation does not affect the primary 3508 FIA trend.
The weak character of D7 meant that the 3508 FIA was
maintained in any andalusite that did not nucleate on
Fig. 4. Location of samples TC1365 and TC1365i from adjacent thin pelitic bands in the Tommy Creek Block of the Mount Isa Province of Queensland,
Australia.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 239
-
8/3/2019 Bell 2006 J Struct Geol
5/17
previously formed porphyroblasts. Very locally, andalusite
has replaced staurolite (Fig. 6c and d) and in these locations
could preserve the staurolite FIA trend.
In layer C, where garnet porphyroblasts first grew early
during the development of sub-vertical D4, the 3508 FIA trend
is controlled by the strike of sub-vertical S4 (Fig. 8a and c;see
also Fig. 2) There was no rotation of S4 preserved in the D5strain shadows of the porphyroblasts (Fig. 7d) during minor
rim growth that accompanied the development of sub-
horizontal S5 and so the FIA defined by L45 remained at
3508. Staurolite grew predominantly during the development
of sub-horizontal D5, which as described in point (2) above,
caused rotation of L45 towards L55 prior to porphyroblast
nucleation and resulted in the 458-trending FIA (Fig. 8d).These observations from adjacent layers reveal that
variation in FIA trend is possible during a single sub-
horizontal foliation producing deformation event for a first
phase of porphyroblast growth in some layers versus a
second or later phase in others. For example, the different
FIA trends in garnet porphyroblasts that grew during D5 in
layers A and B can be readily interpreted via the effects that
the presence of a porphyroblast has on the development and
preservation of foliations in the matrix against it. Further-
more, the parallelism of FIA trends in garnet and andalusite
porphyroblasts that grew for the first time during a sub-
vertical foliation producing deformation event, with those in
garnet in layer A overgrowing earlier formed cores, has
resulted from the only variation that is possible due to
heterogeneous strain being in the vertical plane around a
steep stretching lineation (Fig. 2).
Thus the variation in FIA trend from phase to phase and
layer to layer can be readily interpreted when the timing of
porphyroblast development is examined with regards to
whether the foliation that accompanied growth developed
sub-horizontally or sub-vertically. Where foliations preserved
in porphyroblasts are successively sub-vertical and sub-
horizontal, the FIA trends are controlled by the strike of sub-
vertical foliations as shown in Fig. 8a. Once a porphyroblast
has formed, sub-vertical foliations that formed against it arepreserved in its strain shadow from the effects of rotation due
to subsequent deformation or reactivation of bedding. Any
foliation pre-dating porphyroblast growth canbe rotatedby the
effects of prior deformation, or locally by that accompanying
porphyroblast nucleation.
3. Inclusion trail geometries in 3-D
A significant quantity of data has been published on the
measurement and analysis of FIA orientations with the
Fig. 5. Spiral-shaped inclusion trails (a) and line diagram (b) of a garnet porphyroblast from layer A in sample TC1365i. Vertical section strikes 908 with single
barbed arrow showing strike and way up. The relics of four foliations are preserved with the fifth (S 6) being continuous with the matrix foliation. The
overgrowth of S6 took place during a weak crenulation event with a sub-horizontal axial plane (S7). Note the truncational character of S4, S5 and S6. Plane
polarized light.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252240
-
8/3/2019 Bell 2006 J Struct Geol
6/17
majority of this work carried out using the approach
established by Hayward (1990) of cutting multiple vertical
thin sections around the compass. This method averages the
asymmetry of the inclusion trails in a large population of
porphyroblasts from a single rock sample using several
differently oriented thin sections. Reconciliation of the
structures observed in differently oriented thin sections into
a complete 3-D presentation of the inclusion trail geometry
is generally impractical because each porphyroblast is cut
somewhere between its centre and edge. This does not
change the inclusion trail asymmetry for one stage of growth
but it may change the shape (e.g. compare Fig. 9ac). In
particular, it may reveal curvature of inclusion trails, in some
porphyroblast cores, which pass towards the horizontal
Fig. 6. (a) and (b) Shows straight to slightly sigmoidal shaped inclusion trails in garnet porphyroblasts that grew during the development of a crenulation with asub-horizontal axial plane S5 in layer B in sample TC1365i. Vertical section strikes 1208 with single barbed arrow showing strike and way up. Plane polarized
light. (c) and (d) Shows a staurolite porphyroblast that grew during the development of a crenulation with a sub-horizontal axial plane S 5 in layer B in sample
TC1365i. The staurolite was partially replaced by andalusite in a younger steep event S6. Partial evidence for this is the remains of S4, reactivated during D5,
preserved in the andalusite extremities. Vertical section strikes 1208 with single barbed arrow showing strike and way up. Plane polarized light.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 241
-
8/3/2019 Bell 2006 J Struct Geol
7/17
(Fig. 9a and b) or vertical; such curvature has significance (see
Discussion).
Numerous studies have attempted to predict the
geometry of inclusion trails based on the process by
which they form (e.g. Schoneveld, 1977; Bell and Johnson,
1989; Masuda and Mochizuki, 1989; Williams and Jiang,
1999), but relatively little data has been presented on the
3-D geometry of inclusion trails preserved in real rocks. A
doubly curving non-cylindrical geometry of simple
sigmoidal inclusion trails has been described by several
studies (e.g. Johnson, 1993) utilising a serial sectioning
approach in which the inclusion trails were reconstructed
from tracings of several parallel thin sections made through
a single large porphyroblast. Marschallinger (1998)
employed a precision serial lapping and scanning technique
to aid in similar reconstructions.
The advent of high-resolution X-ray CT has permitted
the actual 3-D geometry of curved inclusion trails in garnet
Fig. 7. (a,b) Staurolite grown in flat D5 event in layer C in sample TC1365. Vertical section strikes 908 with single barbed arrow showing strike and way up.
Plane polarized light. (c,d) Bulk of garnet overgrew S3 in steep D4 event in layer C in sample TC1365 (vertical S4 intensifies at lower RH garnet rim). Upperand lower garnet rims grew in D5. Vertical section strikes 908with single barbed arrow showing strike and way up. Plane polarized light.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252242
-
8/3/2019 Bell 2006 J Struct Geol
8/17
porphyroblasts to be studied in detail for the first time (Ikeda
et al., 2002; Huddlestone-Holmes and Ketcham, in press).
The combination of attenuation data with thin section and
microprobe data allows inclusion phases to be distinguished
and separated. The 3-D model of a single garnet
porphyroblast produced by Huddlestone-Holmes and
Ketcham (in press) provides the actual geometry of a
doubly curving single sigmoidal quartz-rich inclusion trail
that runs through the centre of the porphyroblast, paralleled
by numerous ilmenite inclusions. Their model serves as the
template for this study as the geometry of inclusion trails
preserved throughout this porphyroblast can be examinedand computer derived 2-D cuts made through it in any
orientation to quantify and thus understand the significance
of inclusion trail cut effects through a single porphyroblast.
3.1. Establishing a model porphyroblast
The 3-D inclusion trail geometry, produced from the
high-resolution X-ray CT by Huddlestone-Holmes and
Ketcham (in press) on a sample of Cram Hill Formation
from south-eastern Vermont, is shown in Fig. 10a. This
representation was imported into and manipulated in the
3-D modelling, animation and rendering package Discreet
3ds Maxe. This software allows individual phases (garnet,
quartz, ilmenite, etc.) or even individual inclusions to be
treated and managed separately. It is possible to turn
individual parts of the model on or off to highlight aspects of
interest. We stripped away garnet and crack-fill material to
reveal the central quartz inclusion trail (light coloured)
along with higher attenuation opaques (ilmenite) inclusions
(Fig. 10b). For clarity, we removed isolated quartz
inclusions around the periphery of the garnet because this
study is concerned with structures preserved during the first
phase of porphyroblast growth and these inclusions were
overgrown very late. The central quartz inclusion trails were
traced in three dimensions by adjusting the vertex points ofa flat, box-shaped primitive object in 3ds Max. Adjustments
were made so that the resultant continuous inclusion trail
passes through the centres of the quartz inclusions
(Fig. 10c), resulting in a continuous representation of the
inclusion trail geometry. Removing the central quartz trails
reveals that the dark ilmenite inclusions are parallel to the
central continuous trail in Fig. 10d. Additional trails were
made by extrapolation of the central trail, using the ilmenite
inclusions as a guide and adjusting as necessary to create a
satisfactory fit (Fig. 10e). We then replaced the garnet
around the periphery that was established by the original
X-ray CT to produce the final model (Fig. 10f). The
inclusion trails in this model offer the advantage that they
contain no breaks or gaps. A section cut through the garnet
at any orientation or position will always intersect the trails
and produce a continuous trace.
3.2. Slicing the model
The trend of the FIA is usually found by looking for the
flip in inclusion trail asymmetry in a set of radial vertical
thin sections. Fig. 11 illustrates the inclusion trail
geometries encountered in a set of such radial sections
made through the centre of the model porphyroblast. In
Fig. 8. (a) Shows how FIA trend is controlled by the strike of the sub-
vertical foliation. (b)(e) Skeletal inclusion trail geometries drawn as
vertical cross-sections for each porphyroblast phase in each layer. The
foliations represent those preserved within the porphyroblasts and can be
correlated from layer to layer. (b) The FIA defined by L34, L45, L
56 and L
67 in
the garnet from layer A remain consistently oriented throughout
porphyroblast growth at 3508. (c) The FIA defined by L34 in garnet from
layer C formed during the development of a sub-vertical foliation and trendat 3508. (d) The FIA defined by the L45 intersection lineation were rotated
towards L55 during the development of sub-horizontal S5 in garnet and
staurolite in layers B and C, respectively, and trend at 458. (e) The FIA
defined by the L56 intersection lineation in andalusite in layer B may have
been rotated towards the steeply plunging L66 during the development of the
sub-vertical S6 but the trend remains at 3508.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 243
-
8/3/2019 Bell 2006 J Struct Geol
9/17
reality, we need only cut sections at 0, 30, 60, 90, 120 and
1508 to get this information. Sections at 180, 210, 240,
270, 300 and 3308 can be examined simply by reversing
the initial set of six sections. It can be seen that the flip in
asymmetry from Z to S shaped occurs at, or very close
to, 08 (0008 or 1808 section). Consequently, the FIA for
this sample is oriented approximately NS. The trend of
the FIA can be further constrained by cutting additional
sections close to 0008 (Fig. 12). These additional sections
confirm that the flip in asymmetry occurs very close (G
58) to 08.
Fig. 11 shows that the sections cut perpendicular to the
FIA (0908 and 2708) display the least apparent curvature ofthe trails. These sections, passing through the centre of the
porphyroblast, provide true profiles of the inclusion trail
geometry. Sections cut in the remaining orientations display
an increase in the apparent curvature, appearing more spiral
like, culminating in the closed loop structures of sections cut
parallel to the FIA (0008 and 1808). Structures observed in
sections cut very close to the axis of curvature (e.g. the 010 8,
0208 and 3508 sections of Fig. 12) exhibit almost 1808 of
apparent rotation from the centre to the margin of the
porphyroblast. A comparison of sections made through the
trails of the schematic model and the original model of
Huddlestone-Holmes and Ketcham (in press) is presented in
Fig. 12. The sections through the schematic model clearly
reflect the geometry represented by the central quartz
inclusions of the original model but the schematic model has
the advantage that it is unaffected by the paucity of
inclusions, particularly opaques/ilmenite.
3.3. The cut effect
All sections pass through the centre of the porphyroblast
in Fig. 11. However, it is unlikely that any particular
porphyroblast will be perfectly bisected by a thin section.
Consequently, the inclusion trails observed usually vary
from central cuts as shown for a set of radial thin sections in
Fig. 13 (the numbers in parenthesis are those section
orientations that can be obtained simply by flipping the
original section over; 2108 is equivalent to 0308 viewed
from the opposite direction). Significantly, the flip in
asymmetry occurs very close to 08 regardless of whether
the sections utilised pass through the centre or closer to the
edge of the porphyroblast for one phase of porphyroblast
growth. That is, the asymmetry method for determination of
the axis of curvature is unaffected by cut effects. However,
the doubly curving, non-cylindrical nature of the inclusion
Fig. 9. Shows how reconciliation of the structures observed in differently located thin sections into a complete 3-D presentation of the inclusion trail geometry
is impractical because thin sections commonly cut each porphyroblast represented somewhere between its centre and edge.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252244
-
8/3/2019 Bell 2006 J Struct Geol
10/17
Fig. 10. (a) The 3-D inclusion trail geometry, produced from the high-resolution X-ray CT by Huddlestone-Holmes and Ketcham (in press) on a sample of
Cram Hill Formation from south-eastern Vermont (sample V209 in Bell and Hickey, 1997). (b) Shows garnet and crack-fill material stripped away to reveal the
central quartz inclusion trail (light coloured) along with higher attenuation opaques (ilmenite) inclusions. (c) 3-D inclusion surface drawn to pass through the
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 245
-
8/3/2019 Bell 2006 J Struct Geol
11/17
trails results in a significant decrease in the apparent
curvature of the inclusion trails from the centre to the
outer margin of the porphyroblast. This is especially
apparent in sections cut nearly orthogonal to the FIA
( e. g. a t 0 908). While this can potentially m ake
determination of asymmetry difficult, these effects can
be overcome by examining a large enough population of
porphyroblasts.
Another aspect to cut effects is the orientation of a
section relative to the FIA. Structures observed in sections
cut within 308 of the axis of curvature (Fig. 11 and the 0108,
0208 and 3508 sections of Fig. 12) exhibit almost 1808 of
apparent rotation from the centre to the margin of the
porphyroblast and would be called spiral inclusion trails as
opposed to sigmoidal if viewed in isolation through cutting
just one or two thin sections from this sample. Different sets
of FIAs can show a large range of trends relative to the
matrix foliation (e.g. Ham and Bell, 2004). Therefore, thin
sections cut perpendicular to the matrix foliation and
perpendicular or parallel to a lineation, may cut through a
porphyroblast where the FIA lies at 308 or less to the section
plane. Such thin sections may show perfect spiral shaped
inclusion trails in porphyroblasts where the actual geometry
is a sigmoid (e.g. Fig. 11) and would be interpreted by many
Fig. 11. Shows the inclusion trail geometries encountered in a set of radial sections made through the centre of the 3-D porphyroblast constructed in Fig. 10. In
reality, we need only cut sections at 0, 30, 60, 90, 120 and 1508 to get this information. Sections cut perpendicular to the FIA (0908 and 2708) display the least
apparent curvature of the trails. These sections, passing through the centre of the porphyroblast and cut perpendicular to the FIA can be considered true profiles
of the inclusion trails.
centre of the quartz inclusions. (d) The dark ilmenite inclusions are parallel to the central surface drawn with the quartz inclusions removed. (e) Additional
inclusion surfaces were made by extrapolation of the central trail, using the ilmenite inclusions as a guide and adjusting to create a satisfactory fit. (f) Shows
garnet, as established from the original X-ray CT, resulting in the final model.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252246
-
8/3/2019 Bell 2006 J Struct Geol
12/17
today as a product of intense local shear zone development
(e.g. Williams and Jiang, 1999).
3.4. Variation in FIA trend through the model
Fig. 14a shows three horizontal cuts through the
inclusion trail geometry. The variation in the orientation
of the intersection of the inclusion trails with the horizontal
cut that passes through the porphyroblast centre potentially
defines the maximum variation in trend of a FIA that could
be recorded by the growth of a single porphyroblast that
grew in this event. The intersection of this foliation with the
horizontal cut varies through a maximum of 608. Hud-
dlestone-Holmes and Ketcham (unpublished data) measured
the FIAs separately in 58 porphyroblasts using X-ray CT on
five cylinders cored from different portions of this sample
and found a maximum variation of 308 around the mean FIA
trend giving a total variation of 608. Therefore, the questionto be resolved is how much of this variation predated or was
synchronous with porphyroblast growth. This is particularly
pertinent with this sample because, as mentioned above, it
contains evidence of a foliation inflection through the
vertical that strongly suggests the foliation either initially
had a sub-horizontal orientation (Fig. 14b(i)) or was a
reactivated oblique moderately east dipping foliation sub-
parallel to compositional layering (Fig. 14b(ii)) or was a
pre-existing crenulation about a steep axial plane
(Fig. 14b(iii)). This foliation in Fig. 14b(i) was initially
rotated by a deformation with a sub-vertical axial plane to
the steep east dip or started in this orientation prior to thedeformation with the sub-horizontal axial plane that
accompanied porphyroblast growth. It was then overprinted
by the crenulation with the sub-horizontal axial plane that
accompanied porphyroblast growth. If the porphyroblast did
not overgrow the centre of a partially decrenulated pre-
existing hinge (Fig. 14c(i)), but instead was centred above
that as shown in Fig. 14c(ii), then the resulting foliation
through the centre of the porphyroblast would look like
Fig. 14a(ii) with the actual geometry that should have been
preserved in the porphyroblast centre looking like that in
Fig. 14a(i). Therefore, we conclude that the bulk of the
Fig. 12. The trend of the FIA can be further constrained by cutting
additional sections close to 0008. Sections made through the schematic
inclusion trails are compared with sections made through the original model
ofHuddlestone-Holmes and Ketcham (in press).
Fig. 13. Shows the range of potential inclusion trail patterns made by taking
sections through the schematic model porphyroblast at different orien-
tations and at various distances from the centre. It is statistically unlikely
that a particular porphyroblast will be perfectly bisected by a thin section.
The inclusion trails observed are more commonly variations on these
perfect delineations, as illustrated here.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 247
-
8/3/2019 Bell 2006 J Struct Geol
13/17
variation occurred prior to development of the sub-
horizontal foliation in this sample.
4. Discussion
4.1. The early growth of porphyroblasts relative todeformation
Porphyroblast growth during regional metamorphism
commences very early during a deformation event (Bell
et al., 2003) and ceases once a differentiated crenulation
cleavage begins to form in the immediate vicinity (Fig. 15;
Bell and Hayward, 1991). Nucleation of a porphyroblastic
phase for the first time at a particular site requires irregular
partitioning of the deformation into progressive shearing
and shortening components on the scale of a porphyroblast
(e.g. Fig. 15a), which generally involves relatively coaxial
strain to enhance microfracture along pre-existing foliations
(e.g. Bell et al., 2004). This means that the first phase of
porphyroblast growth generally preserves simple trails that
commonly have a very similar orientation across a thin
section (Ramsay, 1962), fold or region (Fyson, 1980;
Aerden, 2004). However, deformation may locally com-
mence and be distributed very homogeneously on a scale
much larger than a porphyroblast. For example, if
reactivation of the pre-existing foliation is possible or
becomes possible, then the foliation can rotate without
crenulations developing at the scale of a porphyroblast (e.g.
Bell et al., 2003). If this occurs, porphyroblasts will not
nucleate unless the deformation begins to partition into
progressive shearing and shortening components at the scaleof a porphyroblast. Consequently, the foliation preserved as
inclusion trails within such porphyroblasts could vary
significantly from its orientation prior to the commencement
of that deformation event. However, it is significant that in
many examples this is not the case.
Foliations that have been folded several times before
porphyroblast growth begins should show a large range of
variation in orientation when preserved as inclusion trails. It
is intriguing how uncommon this appears to be at the scale
of several outcrops (Steinhardt, 1989; Aerden, 1995), shear
Fig. 14. (a) Shows the inclusion trail geometry in three different horizontal
cuts through the porphyroblast. (b) The inflection through the vertical
shown in (a) can be explained by overprinting during a deformation that
develops a sub-horizontal axial plane shown in a vertical cross-section
through (i) a foliation that was initially flat lying but which had been
overprinted by an earlier steep event, (ii) a foliation that was reactivated and
dipping right, or (iii) a pre-existing crenulation. (c) Shows (i) a
porphyroblast nucleating in the centre of inflection described in (b) and
(ii) a porphyroblast nucleating in the upper part of the same inflection.
Fig. 15. Shows how porphyroblast growth that has commencedvery early during a deformation event ((a)(c)) ceases once a differentiatedcrenulation cleavage
begins to form in the immediate vicinity ((d)(f)). Also shows how nucleation of a porphyroblastic phase for the first time at a particular site requires
partitioning of the deformation into progressive shearing and shortening components on the scale of a porphyroblast through that location.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252248
-
8/3/2019 Bell 2006 J Struct Geol
14/17
zone (Jung et al., 1999), or region (Fyson, 1980). Possible
reasons for this are:
1. Such variation is less interesting than consistency and
has not been recorded; consequently, it is less well
represented in the literature.
2. Foliations lying parallel to bedding are crenulated at a
coarser scale than recently developed axial planar ones
causing porphyroblasts to preferentially nucleate in the
latter.
A significant factor that would promote the situation
described in point 2 is that compositional layering always
will reactivate if the geometry is appropriate and this
reduces the number of sites available for porphyroblast
growth (Bell et al., 2003).
4.2. Porphyroblast nucleation and initial growth
As a result of the early commencement of porphyroblast
growth relative to deformation in the vicinity, the most
common inclusion trail geometry preserved is one of
straight inclusion trails that are slightly curved at the
porphyroblast edges (Fig. 15c). Porphyroblast growthceases once inclusion trail curvature becomes significant
because the progressive shearing causing foliation curvature
also causes a differentiated crenulation cleavage to begin to
develop. The dissolution and solution transfer associated
with this process stops further porphyroblast growth
(Fig. 15df; Bell and Hayward, 1991). Consequently,
porphyroblasts will generally not nucleate over a foliation
lying at a low angle to that which is newly developing assuch foliations will reactivate antithetically as shown in
Fig. 1 (e.g. Bell et al., 2003) or be reused synthetically (e.g.
Davis and Forde, 1994). The next most common inclusion
trail geometry is one of sigmoidal inclusion trails that are
slightly more curved at the porphyroblast edges than
towards the core. Such trails commonly result from the
overprint of a pre-existing crenulation hinge by a younger
event as shown in Fig. 14b(iii). Their development is very
similar to that just described except that relics of an early
crenulation are present providing more curvature of the
inclusion trails.
4.3. Significance of foliation curvature towards the vertical
or horizontal
Multiple thin sectioning of samples commonly reveals,
in some porphyroblasts, inflections towards the vertical or
horizontal for the earliest observable portion of the inclusion
trail as shown in Fig. 16. These inflections have significance
if inclusion trails result from the development of successive
sub-vertical and sub-horizontal foliations as first argued by
Bell and Johnson (1989). They can form in one of three
ways.
1. They reflect the remains of an older crenulation that
predates porphyroblast growth as shown in Fig. 14b(iii).
2. They reflect the remains of a reactivated foliation that
lies oblique to the horizontal or vertical (Fig. 14b(ii)).
3. They result from relatively coaxial deformation occur-
ring early during the deformation that occurred prior to
the first phase of porphyroblast growth (Fig. 15ac; Bell
and Bruce, unpublished data).
If they result from nucleation on an older crenulation,
then one would expect that X-ray CT scans of 3-D inclusion
trail geometries from many porphyroblasts in the same
sample, would reveal such geometries in a lot of them. Thiscould also apply if the foliation was the remains of an
Fig. 16. Skeletal diagrams of vertical cross-sections through spiral-shaped
inclusion trails in porphyroblasts. Curvature towards horizontal of the
central inclusion trail in (a) suggests the presence of an earlier formed sub-
horizontal foliation as shown in (b). Curvature towards vertical of the
central inclusion trail in (c) suggests the presence of an earlier formed sub-
vertical foliation as shown in (d).
Fig. 17. Equal area rose diagrams showing the pitch of truncational discontinuities preserved as inclusion trails measured from between 8 and 12 vertical thin
sections with different strikes per sample (from Gavin, 2004).
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 249
-
8/3/2019 Bell 2006 J Struct Geol
15/17
oblique reactivated one. Indeed this is the situation recorded
by Huddlestone-Holmes and Ketcham (unpublished data).
If, however, they formed by early coaxial deformation one
would expect that the variation in geometry from
porphyroblast to porphyroblast would be very large (e.g.
Bell and Bruce, unpublished data). If such changes occurred
at a much larger scale than a porphyroblast then noporphyroblast would tend to nucleate. However, if
repartitioning of the deformation occurred at a finer scale
through the outcrop, any porphyroblasts that nucleated
would overgrow the previously rotated foliation and perhaps
have a similar geometry to that described for points 1 and 2
above. At a larger scale this might not be the case and the
inclusion trail geometry could vary significantly from
porphyroblast to porphyroblast.
4.4. Curvature of the intersection or inflexion axis within a
porphyroblast
When a porphyroblast first nucleates the FIA could show
considerable curvature in 3-D (Fig. 10) causing a change in
trend in a horizontal plane in Fig. 14a(i)(iii) and plunge in a
vertical plane in Fig. 14b. This variation can be recorded
using X-ray CT methods within individual porphyroblasts
as shown in Fig. 10. However, the method of measurement
used to record the FIA and described by Hayward (1990)
records the FIA for a sample and not for an individual
porphyroblast. The effects of such primary variation within
a porphyroblast are eliminated.
4.5. Apparent spiral shaped inclusion trails due to sectionorientation
The 3-D geometry of inclusion trails in the garnet
porphyroblast revealed by X-ray CT shows that some
section orientations in space contain a spiral shaped
inclusion trail geometry even though the inclusion trails
are sigmoidal in 3-D (e.g. the 0308 and 1508 sections in
Fig. 11). This possibility will be recognized by anyone who
works extensively with multiple thin sections around the
compass but will not be known by most metamorphic and
structural geologists because they have not done this type of
work. Consequently, where thin sections are cut either
randomly or relative to the matrix microstructures, the
porphyroblasts can contain spiral shaped inclusion trails that
may even appear to be continuous with the matrix foliation
(Cihan, 2004), when in fact only sigmoidal ones are present
in 3-D. This could dramatically influence interpretation of
the structural history of the region, depending on whether
one interpreted that the inclusion trails had formed by
porphyroblast rotation or non-rotation. Wherever spiral
shaped inclusion trails are seen in thin section showing a
maximum of 1808 of curvature, extra thin sections in other
orientations should be cut to check whether true spirals or
just sigmoidal trails are present.
Fig. 18. Stereonets of FIA plunges and histograms of angle of plunge data
from (a) and (b) the Spring Hill syncline (Bell and Hickey, 1997), (c) and
(d) the Bolton syncline (Hickey and Bell, 1999), (e) and (f) the Wepawaug
fold (Bell and Newman, unpublished data), (g) and (h) the European Alps
(Bell and Wang, 1999) and (i) and (j) these four regions all combined into
one plot. The plunges are pre-dominantly sub-horizontal ranging up to 308.
The three steep plunges recorded in the Alps all occur in samples where the
porphyroblasts grew for the first time during the development of a sub-
vertical foliation.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252250
-
8/3/2019 Bell 2006 J Struct Geol
16/17
4.6. Geometry of foliation development against a
porphyroblast
Once a porphyroblast has formed it provides a rigid or
competent object against which the shearing component of
the deformation in the matrix tends to nucleate soon after
the commencement of deformation. The large amount of
quantitative data available on the orientation of foliations
forming after an initial stage of porphyroblast growth
suggests that they form predominantly sub-vertically or sub-
horizontally (Fig. 17, modified from Gavin (2004); see also
fig. 10 in Hayward, 1992). This is confirmed by the shallow
plunges of FIAs in the terrains where these have been
measured (Fig. 18 from data in fig. 9d of Bell and Hickey
(1997); table 1 of Bell and Wang (1999); fig. 9 of Hickey
and Bell (1999); and Bell and Newman, unpublished data).
FIAs resulting from successive steep foliations would have
plunges ranging from mainly very steep to locally very
gentle (Fig. 19), which is not what is observed where
porphyroblasts have grown for a second or subsequent time.
The only way to achieve plunges less than 308 during every
phase of porphyroblast growth after the first is for the FIAs
to result from the intersection of successively preserved sub-
vertical and sub-horizontal foliations. Such sub-horizontal
plunges are independent of whether a sub-horizontal
foliation overprints a sub-vertical foliation or a sub-vertical
foliation overprints a sub-horizontal foliation.
4.7. The FIA method versus the cut effect
The FIA method uses the asymmetry of inclusion trails in
all porphyroblasts present in a thin section. Therefore, it
uses all cuts through a porphyroblast from the core to the
rim, in at least eight different orientations around the
compass. Conceptually, all slices through one period ofgrowth of the porphyroblast should preserve the asymmetry
and the FIA method should be independent of any cut effect,
but it is clearly important to test this. Fig. 13 shows that this
is the case. Therefore, the FIA method provides an excellent
method for measuring foliation inflection/intersection axes
preserved within the porphyroblasts within a sample that is
independent of any cut effect on the inclusion trail geometry
through a porphyroblast.
4.8. Consistency of FIA trends
Where FIAs have been measured regionally they
maintain remarkably consistent successions of trends for
large distances (e.g. Aerden, 2004; Bell et al., 2004). In a
range of locations, individual FIAs result from successions
of up to seven foliations intersecting in the one axis
requiring that directions of relative bulk shortening remain
constant for significant periods of time. Bell and Welch
(2002) suggest that these periods are around 1030 million
years in length. Layer A in sample TC1365i contains garnet
porphyroblasts preserving evidence for five foliations plus a
crenulation axial plane in the matrix that intersect in the one
axis at 3508. Variation from this trend only occurs in
samples where there is one phase of growth that occurred
during the development of a horizontal foliation. Suchsamples can be readily eliminated from a regional analysis if
anomalies relative to multiple FIA samples are recorded.
Acknowledgements
We acknowledge the support of the ARC for this
research and the excellent facilities provided by the School
of Earth Sciences at James Cook University for microstruc-
tural research. We acknowledge James Lally for collecting
samples TC1365 and TC1365i and bringing them to our
attention.
References
Aerden, D.G.A.M., 1995. Porphyroblast non-rotation during crustal
extension in the Variscan LysCaillaouas Massif, Pyrenees. Journal
of Structural Geology 17, 709725.
Aerden, D.G.A.M., 2004. Correlating deformation in Variscan NW-Iberia
using porphyroblasts; implications for the Ibero-Armorican Arc.
Journal of Structural Geology 26, 177196.
Bell, T.H., Hayward, N., 1991. Episodic metamorphic reactions during
orogenesis: the control of deformation partitioning on reaction sites and
duration. Journal of Metamorphic Geology 9, 619640.
Fig. 19. Stereonet showing how the intersection of a sub-horizontal plane
with any other steeply dipping plane produces a gentle plunge (circles). It
also shows how the intersections of steeply dipping planes have mainly
steep plunges. They only range to gentle plunges when the intersecting
steeply dipping planes are very close in strike, and mainly have opposite
dips.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252 251
-
8/3/2019 Bell 2006 J Struct Geol
17/17
Bell, T.H., Hickey, K.A., 1997. Distribution of pre-folding linear
movement indicators around the Spring Hill Synform, Vermont:
significance for mechanism of folding in this portion of the
Appalachians. Tectonophysics 274, 275294.
Bell, T.H., Hickey, K.A., 1998. Multiple deformations with successive sub-
vertical and sub horizontal axial planes: their impact on geometric
development and significance for mineralization and exploration in the
Mount Isa region. Economic Geology 93, 13691389.
Bell, T.H., Johnson, S.E., 1989. Porphyroblast inclusion trails: the key to
orogenesis. Journal of Metamorphic Geology 7, 279310.
Bell, T.H., Wang, J., 1999. Linear indicators of movement direction versus
foliation intersection axes in porphyroblasts (FIAs) and their
relationship to directions of relative plate motion. Earth Science
Frontiers 6, 3146.
Bell, T.H., Welch, P.W., 2002. Prolonged Acadian orogenesis: revelations
from foliation intersection axis (FIA) controlled monazite dating of
foliations in porphyroblasts and matrix. American Journal of Science
302, 549581.
Bell, T.H., Ham, A.P., Hickey, K.A., 2003. Early formed regional antiforms
and synforms that fold younger matrix schistosities: their effect on sites
of mineral growth. Tectonophysics 367, 253278.
Bell, T.H., Ham, A.P., Kim, H.S., 2004. Partitioning of deformation along
an orogen and its effects on porphyroblast growth during orogenesis.
Journal of Structural Geology 26, 825845.
Cihan, M., 2004. The drawbacks of sectioning rocks relative to fabric
orientations in the matrix: A case study from the Robertson River
Metamorphics (Northern Queensland, Australia). Journal of Structural
Geology 26, 21572174.
Davis, B.K., Forde, A., 1994. Regional slaty cleavage formation andfold axis
rotation by reuse and reactivation of pre-existing foliations: the Fiery
Creek Slate Belt, north Queensland. Tectonophysics 230, 161179.
Fyson, W.K., 1980. Fold fabrics and emplacement of an Archean granitoid
pluton, Cleft Lake, Northwest Territories. Canadian Journal of Earth
Sciences 17, 325332.
Gavin, B., 2004. The microstructural and metamorphic history preserved
within garnet porphyroblasts from southern Vermont and northwestern
Massachusetts. Unpublished PhD thesis, James Cook University,
184pp.Ham, A.P., Bell, T.H., 2004. Recycling of foliations during folding. Journal
of Structural Geology 26, 19892009.
Hayward, N., 1990. Determination of early fold axis orientations within
multiply deformed rocks using porphyroblasts. Tectonophysics 179,
353369.
Hayward, N., 1992. Microstructuralanalysisof the classicsnowballgarnets of
southeast Vermont. Evidence for non-rotation. Journal of Metamorphic
Geology 10, 567587.
Hickey, K., Bell, T.H., 1999. Behaviour of rigid objects during deformation
and metamorphism. a test using schists from the Bolton Synform,
Connecticut. Journal of Metamorphic Geology 17, 211228.
Huddlestone-Holmes, C.R., Ketcham, R.A., 2005. Getting the inside story:
using computed X-ray tomography to study inclusion trails in garnetporphyroblasts. American Mineralogist, in press.
Ikeda, T., Shimobayashi, N., Wallis, S.R., Tsuchiyama, A., 2002.
Crystallographic orientation, chemical composition and three-dimen-
sional geometry of sigmoidal garnet: evidence for rotation. Journal of
Structural Geology 24, 16331646.
Johnson, S.E., 1993. Unraveling the spirals: a serial thin section study and
three-dimensional computer-aided reconstruction of spiral-shaped
inclusion trails in garnet porphyroblasts. Journal of Metamorphic
Geology 11, 621634.
Jung, W.S., Ree, J.H.,Park, Y., 1999. Non-rotation of garnet porphyroblasts
and 3-D inclusion trails dataan example from the Imjingang Belt,
South Korea. Tectonophysics 307, 381395.
Lally, J.H., 1997. The structural history of the central eastern fold belt,
Mount Isa Inlier, Northwest Queensland, Australia. Unpublished PhD
thesis, James Cook University, 214pp.Marschallinger, R., 1998. Three-dimensional reconstruction and modelling
of microstructures and microchemistry in geological materials.
Scanning 20, 6573.
Masuda, T., Mochizuki, S., 1989. Development of snowball structure:
numerical simulation of inclusion trails during synkinematic porphyr-
oblast growth in metamorphic rocks. Tectonophysics 170, 141150.
Ramsay, J.G., 1962. The geometry and mechanics of formation of similar
type folds. Journal of Geology 70, 309327.
Sayab, M., 2005. Microstructural evidence for NS shortening in the Mount
Isa Inlier (NW Queensland, Australia): the preservation of early WE-
trending foliations in porphyroblasts revealed by independent 3D
measurement techniques. Journal of Structural Geology 27, 14451468.
Schoneveld, C., 1977. A studyof some typical inclusion patterns in strongly
paracrystalline rotated garnets. Tectonophysics 39, 453471.
Steinhardt, C.K., 1989. Lack of porphyroblast rotation in non-coaxially
deformed schists from Petrel Cove, South Australia, and its
implications. Tectonophysics 158, 127144.
Williams, P.F., Jiang, D., 1999. Rotating garnets. Journal of Metamorphic
Geology 17, 367378.
T.H. Bell, M.D. Bruce / Journal of Structural Geology 28 (2006) 236252252